As I mentioned in a previous post, one of the main goals I had in mind for version 0.9.8 of Project Darkstar was establishing an application deployment mechanism that was both simple and consistent. A cornerstone feature of any application container, at least from an IT and system administrator's standpoint, is a clearcut and highly predictable structure for application deliverables. This streamlines maintenance, keeping the costs down and contributing to just one of the value-adds of the technology. Personally, I think that we accomplished this with this release.
There is a flip side to this coin, however. In addition to administrator and deploy-time efficiencies, there are also developer use-cases that require simple, consistent, and automated bring up of a deployed Project Darkstar application. Whether it's for testing, debugging, or one-off temporary deployments, the need for quickly and automatically booting up a Project Darkstar server application during development is obvious. The question is, can we leverage the new distribution structure of 0.9.8, which accomplishes our first goal, to also establish a "best practice blueprint" for our second goal? I think the answer is yes.
The Project Darkstar Maven Plugin does exactly this. Using the standardized structure established in version 0.9.8, the plugin provides the capability to control a Project Darkstar server installation, including deployment, bootup, and shutdown. Is it a silver bullet? Probably not. But I've already ported the wanderer example and Project Snowman to use it with good success. If you're developing a Project Darkstar application, and especially if you're using Maven, I would encourage you to check it out.
No comments :
Post a Comment